All the patents in suit are for improvements to oil-burning furnaces for heating buildings, and they all follow the same general pattern. An electric motor m the service circuit drives a pump which feeds the fuel — oil'—-by a pipe into the furnace. At some point, usually at the outlet of the pipe, the oil is mixed with air and discharged into the furnace in a spray. In the earlier systems a pilot light set this on fire, but in the fully developed art an electric spark was substituted. When the control room gets too cold, a thermostat closes the motor circuit and starts the motor; but, since at times the mixturé does not catch fire, or goes out after it has, the pump still *918keeps working, for the control room remains cold, or becomes so, and the room thermostat does not open the motor circuit. To fill the furnace with the unburnt mixture is dangerous, for it is explosive and may cause serious accidents. All three of the patents in suit were for improvements upon such a system; they were designed to stop the motor when the mixture did not catch, or when it went out. More properly they were for improvements upon what the art had already done to bring about that result. We need describe only one earlier patent, which issued to Sherman & Sheppard on January 20, 1925, upon an application filed on July 8th, 1918, and which was co-pending with Scott’s reissue patent, the application for the original of' which was filed on April 30, 1921.
This patent disclosed an oil furnace, the motor circuit for which contained two switches: one in the “relay, J”, and the “safety switch, K”. The room thermostat was in a circuit in parallel with the motor circuit, which led through a magnet in the “relay, J”. When the control room was cold, the thermostat closed its circuit, and the magnet moved the switch in the relay which closed the motor circuit. The pump would thereafter supply the furnace with the mixture, whether the pilot light was firing the mixture or not; and to prevent this the patentees introduced a third circuit, also in parallel with the motor circuit, which operated the “safety switch, K”. In this were a second, or furnace, thermostat, “controlled by the temperature in the fire-box”, (lines 63-65, page 3), and the “switch, R”, open when the motor was not working. One contact of the “switch, R” was fastened to the expansible end of the “tank, Q”, which a pipe connected with the blower. When the motor began to work, it pumped air into the tank, which as the pressure increased slowly expanded until it closed the “switch, R”. This closed the third circuit and actuated a magnet in it, which opened the “switch, K” in the motor circuit. In this, way the third circuit would always open the motor circuit, and the pump would work only intermittently. To prevent this the furnace thermostat was rigged to close when the furnace was cold, and to open when it was hot; and the “tank, Q” was designed not to close the. “switch, R”, before the thermostat opened. If, however, the mixture did not catch, or went out the pressure in the “tank, Q” would close the “switch, R”; and as the furnace thermostat remained closed, the “safety switch, K” broke the motor circuit.
This was an entirely operable system and was actually used, but it had two defects, or at least two disadvantages: .(1) both thermostat circuits were in parallel with the service circuit, which is always at relatively high voltage; and (2) the “safety switch, K” was normally closed, and had to open to break the motor circuit. The first was bad because high voltages wear out contacts; the second is said to be dangerous because at times a switch will not open for mechanical reasons. Scott’s reissue patent remedied the first defect; his second patent, the second. The reissue patent, itself an improvement over an earlier patent of Scott (No. 1,320,936, applied for May 9, 1919 and issued Nov. 4, 1919) disclosed the following system. When the room thermostat was open — i. e., when the control room was hot — if the “safety switch” was closed, the service current idled through the primary of a transformer set in a shunt. When the room thermostat closed, it closed a circuit passing through the secondary of this transformer, which we will call the “first therm.ostat circuit”, and the current energized a magnet that swung a pivoted armature. In its new position this armature completed a circuit from the service wire through the motor and started it, broke the “first thermostat circuit” and completed another circuit (the “second thermostat circuit”) which also passed through the secondary of the transformer. (The “second thermostat circuit” actuated the spark circuit by means of a second transformer, and so fired the furnace.) In its course it passed through a switch, one contact of which was fixed to the end of a “Bourdon tube”, fed by the fuel pipe. (A “Bourdon tube” is a curved receptacle which straightens out as it fills). As soon as the motor began to-work, the tube began to fill, quite as Sherman & Sheppard’s pump filled the “tank, Q” with air. Its movement broke the “second thermostat circuit”, but, as the tube carried one contact with it, when it had fully straightened out, it could, and did, connect the contact with another contact, and so completed a third circuit, in which were a magnet, the secondary of the original transformer, and a thermostat-in the furnace. This thermostat was open when the furnace was hot, but closed when, it was cold. Thus, if the mixture did not. *919light or went out, the third circuit closed after the Bourdon tube had straightened, and actuated the magnet, which thereupon opened a switch in the motor circuit. The system was then dead until the switch was manually set again. The operation of this system presupposed that the Bourdon tube would not fully open to make contact in the third circuit until the furnace thermostat had opened its contact in that circuit.
The plaintiff urges two points of novelty in this disclosure over Sherman & Sheppard; first, that the furnace thermostat is “mounted in the combustion flue, 2”, (page 2, line 34), while the thermostat in Sherman & Sheppard is on the outside of the furnace; and second, that the subsidiary circuits all derive their current from the secondary of the transformer, and are therefore at lo'w voltage. Except for these features all the claims in suit except 21 read upon Sherman & Sheppard, and the differences in claim 21 are not patentable. The argument is that the “thermostat, P,” of Sherman & Sheppard was not “directly responsive to and regulated by combustion conditions”. The Eighth Circuit held the contrary in Electrol, Inc. v. Merrell & Co., 39 F.2d 873, 877, and there can be no question that it was right. We have already seen that this thermostat was “controlled by the temperature of the fire-box”, not by the pilot light as the plaintiff would have us believe. In figure one it was shown as set on the outside of the furnace on the other side of the burner, D, from the pilot light. The burner mixed the air and oil and sprayed it directly into the furnace; the pilot light set it on fire just as it emerged from the tip of the burner; and it is unbelievable that, so situated it should have been expected to heat the thermostat, even if the words of the specification had not been to the contrary. Against this the plaintiff cites several passages which spoke of the thermostat as depending upon the pilot light. That was quite natural and is readily understandable, for, if the pilot light went out, the “temperature of the fire-box” would certainly fall; just as it would normally rise if it were lighted. Probably the patentees did not think of the possibility that the light, though burning, might not fire the mixture; they were right m not doing so, for it was too remote a probability; but, whatever they thought, what they said provided against just that eventuality, and nobody can follow their instructions who does not make the thermostat depend upon the temperature of the furnace. Besides, surely if it was in the wrong place, experience would cure the defect.
As to the second point, considering the short time between the filing dates of the two applications and the absence of any intervening unsuccessful efforts, we should have great doubt that it was invention merely to make the subsidiary circuits carry a low voltage; if trouble had developed from the use of the service current, the remedy would readily have suggested itself. However, we need not speculate, for the patent to Wales, (No. 1,393,-654) — applied for on February 17, 1921, about two months before the Scott reissue patent — disclosed just such a device. In it the pump was driven by a motor which took its current from a high voltage circuit, open at the “switch, 43” when the room thermostat circuit was itself open. In the thermostat circuit was the secondary of a transformer, actuated by a shunt from the high voltage circuit. When the control room thermostat closed its circuit, it energized a magnet which closed the “switch, 43” and started the motor. In the thermostat circuit was a second thermostat, set so as to be subject to the heat of the pilot light that fired the furnace. If that light went out, the second thermostat broke the thermostat circuit, and by opening the switch in the motor circuit stopped the motor and the pump. This was certainly an embodiment of the low voltage feature introduced by Scott’s reissue, assuming that a patent could in any case have rested upon the improvement. The plaintiff argues that since Wales provided only for pilot light regulation, it was not a complete anticipation. That is true; but it would not be patentable to combine a low.voltage circuit to make and break the motor circuit, with “combustion” regulation, as opposed to combining the same circuit with pilot light regulation.
Scott’s patent No. 1,602,175 was applied for on May 31, 1924, not quite two years after the original of his reissue patent had issued. A shunt from the service wire ran through the primary of a transformer; and the closing of the room thermostat completed a circuit containing the secondary of this transformer, two magnets and a second thermostat. The magnets attracted an armature which tilted a mercury switch that closed two terminals in the motor circuit and started the motor. *920(At the same time a second shunt from the motor circuit actuated a second transformer, whose secondary was part of the spark circuit.) In parallel with the room thermostat circuit was a resistance coil wound about a second thermostat, itself in the room thermostat circuit. As the coil heated, it warped this second, or resistance coil, thermostat so as to break the room thermostat circuit. (The extreme position of this thermostat broke the spark circuit.) When the magnets went dead, the armature would tilt back to its original position, opening the -motor circuit. To prevent this there was in parallel with the room thermostat circuit and including the magnets but not the resistance coil thermostat, a shunt circuit containing a third thermostat set in the furnace. This thermostat exactly reversed the usual order; it opened the shunt when the furnace was cold, and closed it when it was hot. Another switch in the shunt was open when the armature, carrying the mercury switch, was not upon the magnets, but closed, when it was.
The operation was as follows. When the room thermostat closed, it energized the magnets, and by tilting the mercury switch started the motor. The spark should start 'at the same time and the mixture should catch fire. If it did, the furnace thermostat would close and the shunt circuit would remain unbroken, after the room thermostat circuit opened at the resistance coil thermostat. When this phase was esstablished, the pump would operate indefinitely, unless the room thermostat opened and deenergized the magnets. If the mixture did not catch, the furnace thermostat did not close, and the motor stopped when the resistance coil thermostat opened, because the shunt .did not continue to hold the armature. The same was true if the furnace, having once been fired, went out.
In Sherman & Sheppard, and Scott’s first patent, No. 1,320,936, the furnace thermostat opened when the furnace was hot, and closed when it was cold; and so it was in Scott’s reissue patent. Nor did Wales quite reverse this relation, although the defendant asserts that he did. It is true that in Wales’ disclosure the pilot light had to be burning to close the furnace thermostat, without which it would have done no good to close the room thermostat. But the fact that the pilot light was lighted did not inevitably mean that the furnace was afire; nor would it have done so, though the pilot light had been actually set in the furnace. Nevertheless, so far as we can see, (there is no evidence in the record), that was altogether unlikely; besides, if it happened, the mixture could hardly have been very inflammable anyway. Moreover, assuming that it was a useful detail of design, we are very doubtful that it would support a patent. Once it was conceived as desirable, the means to accomplish it were fairly apparent. In the earlier forms the furnace thermostat broke the circuit whose mechanical action in time would trip the switch in the motor circuit; it did so by opening and obviously it must, open when the furnace was hot. But if one wished to reverse this and make the furnace thermostat open when cold, it could not break the circuit that controlled the delayed action device, for it must do that when it was itself hot. One -obvious way, perhaps the only one, as it seems to us, was to put the furnace thermostat in a circuit which would hold the switch in the motor circuit closed, notwithstanding the breaking of the circuit of the delayed action device; and one way to do that was to shunt that device. Nor can we see that the mere idea of a safety switch, which should close when hot, deserved a patent. Certainly it did not, if experience had disclosed defects in the opposite form; if it had not, and the idea merely came out of the blue, we have no assurance that time would not have called it forth; and if time would not, the improvement could scarcely have been of importance. If therefore we-were pressed to a conclusion, we should hesitate to hold that claims two and six were valid. That, however, we will not do, for, along with the Eighth Circuit, we think that, if valid at all, they must be narrowly confined, and that the defendant, does not’infringe them.
If we accept the plaintiff’s application of the terms to the defendant’s installation, the “means for operating and maintaining-combustion” in the claims is the “relay”; just what that covers we are not sure, but we suppose that it includes the secondary of the transformer, together with the-mechanism for tilting the motor switch and that switch itself. If that be true, then,, when applied to the plaintiff’s disclosure,, the same words must cover the magnets, their armature and the mercury switch. So far we will assume that the defendant would infringe. The plaintiff next says. that the “safety device” for controlling the; *921“means for operating”, etc., in the defendant’s system is the circuit of the room thermostat, the secondary and the safety switch heater, but not the latch or the safety switch. In the patent this is the circuit about the magnets, including the “resistance coil thermostat”. Although it is hard to think of the magnets as themselves included in the circuit, let us assume so, and that the defendant would still infringe. The plaintiff then says that the “automatic device” of claim two, which is the same thing as the “electrically-heated thermostatic device” of claim six is represented by the defendant’s safety switch heater, the latch and the switch. It is however not true that the defendant’s safety switch has anything to do with controlling the operation of the relay in the sense that the “resistance coil thermostat” • of the plaintiff’s disclosure controls it. Moreover, even if this were true, the defendant’s “furnace thermostat” which the plaintiff says is the “automatic safety switch” of claim two, and the “automatic switch” of claim six, does not cooperate with the defendant’s safety switch, when closed, to hold the relay when the furnace is hot. The “resistance coil thermostat” always opens, and the cooperation between it and the “furnace thermostat” is by taking over its function. Nothing like that takes place in the defendant’s apparatus; the safety switch never opens unless the whole furnace is to go out of action. Conceivably these claims might serve if the invention were a pioneer; perhaps indeed, if Scott had insisted, he might have been able to get a claim in general terms for any furnace thermostat, open when cold. We cannot know; he did not; and, even though we assume that the claims are valid, they cannot be construed to cover so different a system.
Aldrich’s patent, (No. 1,579,497), is for a further detail in the same kind of machine. It is important in order to save oil and also to some extent for safety, that the pump should stop as soon as possible after the fire dies in the furnace. In the systems before Aldrich, if the thermostat had been moved through a relatively large distance, it would have to move back the same amount before it would make the contact which broke the motor circuit. No changes in its position counted until the circuit was completed; and, since a thermostat ordinarily operates by the differential in expansion of two metals, its movement was proportional to the degree of change in temperature. There might therefore be a de-’ cided drop in heat within the furnace after the fire went out before the pump would stop; and during that period it would keep on filling the furnace with the mixture. Aldrich’s purpose was to make the thermostat immediately sensitive to any change of heat in the furnace, or at least to much smaller changes, so that the pump would stop almost at the same time that the fire went out. This he did by making the thermostat of a metal tube which housed a rod having a smaller coefficient of expansion than the tube. The effect was that as the tube got hot the rod withdrew within it, and came out as it cooled. The rod abutted upon a pivoted lever, and a second rod abutted upon the other side of the lever’s free end, to which a spring pressed it. As the first rod moved under the influence of the heat of the furnace, the second rod would move against the spring in the same direction, and would return as the first rod withdrew. The second rod carried a collar which could slip along it; to the collar was attached one end of a spring, the other end of which was in turn attached to a yoke, and the yoke to a movable contact, which made and broke the circuit that controlled the switch in the motor circuit. On either side of the collar was an abutment through which the second rod moved, but which stopped the collar, thus limiting its movement to the difference between its length and the space between the abutments. When either abutment stopped the collar, it would slip along the second rod while the rod continued to move. This device insured that, no matter what the temperature of the furance, a slight drop would throw off the switch. It would operate equally well though the furnace thermostat were rigged to open when the furnace was cold and to close when it was hot.
We think that the British patent to Jei-del (No. 3818 of 1914) left nothing patentable in Aldrich’s disclosure. Jeidel’s patent was for a thermostat in an electric circuit which a small, and substantially constant, change in heat, would make or break, no matter what the heat at which the. change took place. The thermostat was made of a tube, relatively high in ex-pansibility, housing a rod relatively low. The tube carried one contact of the circuit; a collar upon the rod carried the other and the collar would slip along the *922rod. An insulating piece upon the tube engaged the contact on the collar before tube and collar touched, and the contact on the tube passed around the contact on the collar so that the circuit was broken as soon as the tube expanded. If it continued to expand, the insulating block would touch the contact on the collar and thereafter push it along the rod. If, however, in that position, the tube contracted the distance of the contacts' apart, the circuit would again be made, and thereafter the contact on the tube would push the collar along the rod in the opposite directions. Thus, the distance between the tube and the rod would be confined to a very narrow limit, which was the most that the tube had to move to make or break the circuit, no matter what the temperature of the thermostat.
Aldrich did no more than redesign this thermostat, and insert it in one of the existing types of oil-burning furnace. It is true that the plaintiff criticizes Jeidel’s disclosure on the ground that. it showed no connection between the rod and the tube, without which the thermostat would not operate at all; but nobody could fail to see how it was intended to work, who was not perversely determined not to make it do so. Since Aldrich’s improvement must be judged as though Jeidel’s thermostat was before him, the substitution was not enough. It is true that Sherman & Sheppard filed their application in 1918, but oil furnaces did not become common until the twenties, and Aldrich filed in September, 1923. Thus the wait for his improvement was very short; and it was a minor one when it came, the kind that comes to all new machines as they approach final stability; a refinement that could not be expected upon the first appearance, but that time was likely to develop. It is true that without Jeidel the means chosen might have been patentable, for the device was certainly ingenious; but with that already contrived, we can see nothing that should be held an invention.
We conclude therefore that in the case of Scott’s reissue and of Aldrich, the claims are invalid, and that the defendant does not infringe Scott’s last patent. There is no evidence that any of the three have substantially contributed to the art. It is true that the plaintiff has granted many licenses; but — as is so often the case —they have all covered other patents as well. The inference that any one of the lot was itself an important contribution is therefore far from strong; it is often less troublesome to buy one’s peace from the possessor of many patents, than to enter into a long-drawn and expensive litigation.
Decree reversed; complaint disniissed.